Spontaneous damage to the four bases of DNA is a major cause of the mutations that give rise to cancer. Most of these genetic lesions are corrected by either of two pathways, base-excision DNA repair (BER) or nucleotide-excision DNA repair (NER). BER is primarily responsible for the repair of nucleobases having relatively small, simple changes with respect to the normal nucleobases in DNA, whereas NER repairs a wide variety of bulky nucleobase lesions such as thymine dimers. The key components of BER are DNA glycosylases, professional lesion-hunting enzymes that scan the genome in search of particular kinds of base damage, then catalyze excision of the damaged base from the DNA backbone. NER does not have such specialized lesion-recognition enzymes, but instead employs two proteins, UvrA and UvrB, to search cooperatively for damaged nucleobases of many different kinds, the only common feature being that they are bulky. The long-term goals of our studies are to understand how these enzymes locate their particular kinds of damage amidst the vast expanse of normal DNA, and how they catalyze repair of the damage once having located it. A comprehensive, fundamental understanding of DNA damage recognition and removal represents the solution to a major aspect of the tumorigenesis puzzle. In the proposed studies, we will study base-excision repair proteins of the so-called GO system that are responsible for either direct repair of the highly mutagenic lesion 8-oxoguanine (oxoG) - the hOgg1 enzyme in eukaryotes and MutM in prokaryotes - or repair of the mutagenic adenine in oxoG:A base-pair resulting from mis-replication of unrepaired oxoG lesions, catalyzed by the MutY protein (hMYH in humans). We also propose to investigate lesion recognition by glycosylases that repair a variety of genotoxic methylated adducts in DNA (the AlkA protein in bacteria and Aag in humans). On the NER front, we will study the early events in the pathway leading to the loading of UvrB onto a lesion and recruitment of a UvrB-dependent endonuclease, UvrC, which cleaves the DNA backbone at sites flanking the lesion. Here we outline a broad-based, interdisciplinary approach that employs the use of chemical crosslinking and synthetic, photoactive nucleobase analogs to trap intermediates in the BER and NER repair pathways. To understand dynamic aspects of lesion recognition, we will employ time-resolved X-ray crystallography and single-molecule DNA tracking studies. Together, these studies aim to provide a comprehensive molecular-level framework for understanding two very important but very different strategies for seeking out and destroying genotoxic lesions in DNA.